Genomic divergence between species identifies taxonomic outliers and establishes calibrated molecular clocks

Natalie Marion Allen , Mekala Sundaram, Avril Harder, Jong Yoon Jeon, Gina F. Lamka, Janna R. Willoughby, J. Andrew DeWoody

For decades, molecular clocks that rely on one or a few genes have been employed to quantify evolutionary rates (e.g., anagenesis) or to help date important phylogeographic events (e.g., vicariance). Genome-wide clocks remain largely unexplored, and the degree to which genome-wide divergence scales with time is still unknown in most vertebrates. Here, we expand the concept of the molecular clock to entire genomes by calibrating both nuclear and mitochondrial clocks. We do so in ~700 congeneric tetrapod species pairs using distance estimates from pairwise genome alignments and best estimates of time since divergence. We find that divergence time is the dominant and most consistent predictor of genome-wide divergence, with both mitochondrial and nuclear clocks scaling log-linearly with time. Because our genomic clocks average across entire genomes, they have the ability to smooth over locus-specific variance. The two clocks complement each other temporally, with nuclear clocks being most reliable at deeper divergence times (>5 MYA) and mitochondrial clocks effectively resolving shallower divergence times (<5 MYA). Calibration at narrower taxonomic scales improves accuracy, as illustrated by our Order-level clocks that reflect known biological differences among clades (i.e., life history pace). When applied as diagnostic tools, the clocks recover known cryptic species complexes (e.g., mustached bats) and biological phenomena such as mitochondrial capture (e.g., rock lizards), establishing their utility in flagging candidate species groups for additional evolutionary analyses and/or taxonomic reevaluation.